We have divided this project into three portions. 1) We developed a transgenic mouse model of T-cell leukemia by overexpressing SCL and LMO1 in the thymus. We identified mutations of the Notch1 gene in 70% of these mice, demonstrating that Notch1 activation was a progression event that collaborated with SCL/LMO1. The Notch1 mutations are important in maintaining the malignant phenotype of these cells, as T-cell lines established from SCL/LMO1 mice with Notch1 mutations are sensitive to gamma-secretase inhibitors. We used cre-lox technology to generate mice in which SCL can be activated conditionally. These mice show impaired thymocyte differentiation, with a dramatic increase in immature CD4-/CD8- thymocytes. 2) We have cloned several chromosomal translocations that generate NUP98 fusion proteins. We generated transgenic mice that express NUP98HOXD13 or NUP98TOP1 in the hematopoietic compartment. The NUP98HOXD13 mice develop a highly penetrant myelodysplastic syndrome (MDS) that resembles the human disease in terms of peripheral blood cytopenias, dysplasia, apoptosis, and progression to frank leukemia. We used retroviral tagging experiments to identify genes that collaborate with NUP98-HOXD13 in the process of leukemic transformation, and identified a number of expected (Meis1) collaborating genes, as well as several novel and unexpected collaborators (MN1, Erg, Mir29a). We have shown for the first time that MDS is transplantable, by transplanting bone marrow from NUP98-HOXD13 mice with MDS. In competitive repopulation assays, the NUP98-HOXD13 cells strongly out-compete wild-type cells, indicating that the NUP98-HOXD13 cells have a proliferative and/or homing advantage. We have treated NUP98-HOXD13 mice with a demethylating agent, and shown that 2 of 5 mice had remarkable improvements in their neutrophil and red blood cell count. We generated "knock-in" ES cells that express NUP98-HOXD13 under control of endogenous NUP98 regulatory elements, and have derived immortal, IL-3 dependent hematopoietic precursors from these ES cells. In addition to impaired myeloid and erythroid differentiation, we have recently demonstrated that both B and T lymphoid differentiation are impaired in the NUP98-HOXD13 mice. The T-cells also show a remarkable clonal expansion of cells with a partially rearranged T-cell receptor beta gene. Preliminary experiments suggest that the impaired hematopoietic differentiation is due to up-regulation of Hoxa cluster genes. Finally, we have assayed the myeloid leukemias that develop in the NUP98-HOXD13 mice, and demonstrated that spontaneous Nras, Kras, and Cbl mutations can collaborate with the NUP98-HOXD13 transgene to produce AML. 3) Evaluation of CALM-AF10 as an oncogenic fusion. The CALM-AF10 fusion is caused by a t(10;11) chromosomal translocation seen in a wide spectrum of acute leukemia, but most commonly in T-ALL. We have generated transgenic mice that express the CALM-AF10 fusion in the hematopoietic compartment, 50-70% of these mice develop acute leukemia, predominantly AML. Of note, many of the AML cases have a unique phenotype, with features suggesting partial lymphoid differentiation, such as B220 expression or clonal Igh gene rearrangements. Infection of CALM-AF10 mice with a replication competent retrovirus has identified several common insertion sites; candidate genes located at the common insertions are currently being evaluated to determine if these genes collaborate to produce AML in CALM-AF10 mice. Similar to findings with NUP98-HOXD13 mice, CALM-AF10 mice also show increased expression of Hoxa cluster genes and impaired thymocyte differentiation, suggesting that the cause of leukemia is, at least in part, impaired blood cell differentiation caused by overexpression of Hoxa cluster genes.